polygondragons

Supertonic just killed ElevenLabs. A text-to-speech model that runs entirely on your device. No cloud. No API key. No per-character pricing. 2,700 GitHub stars. 100% open source. MIT licensed. The numbers are wild: → 167x faster than real-time on an M4 Pro → Only 66M parameters → 1,263 chars/sec vs ElevenLabs Flash at 287 → 1,048 chars/sec vs OpenAI TTS-1 at 55 → Runs on a Raspberry Pi. Runs on an e-reader in airplane mode. Reads currency, dates, phone numbers, and technical units correctly without preprocessing. ElevenLabs fails these. OpenAI fails these. Gemini fails these. Supports 11 platforms and 5 languages. Chrome extension turns any webpage into audio in under a second. I've watched on-device models lose to cloud APIs for years. This one doesn't lose. The cloud TTS business just got cooked.

Introducing a 100% free coding agent with DeepSeek v4 Pro Choose any model, all free: - DeepSeek v4 Pro/Flash - Kimi K2.6 - MiniMax M2.7 npm i -g freebuff

隆重推出 CLAUDE MUSIC：估值 9000 亿美元的 @claudeai 正在 YouTube 上直播 LOFI 音乐。

一部手机，干翻整个 3D 漫游行业！ 以前要花六位数才能做的浏览器交互式 3D 漫游，现在一夜之间就能完成 AI 扫描工具正在把普通手机变成全功能 3D 制作工作室

ESTO ES UNA LOCURA La mayoría pasa AÑOS usando apps de idiomas y aun así no logra hablar bien. Claude hizo en 4 semanas lo que Duolingo no pudo arreglar en 4 años. Aquí van los prompts 👇

Sigues pagando 10, 15, 20 suscripciones al mes. Para casi todas existe una alternativa open-source que funciona igual o mejor. Gratis. El problema nunca fue que no existieran. El problema era encontrarlas sin perder tres horas en GitHub. Este sitio te deja buscar proyectos open-source por categoría, lenguaje y popularidad en segundos. Sin cuenta. Sin pagar. Guarda esto antes de renovar otra suscripción.

Someone just built a website where you can find open-source softwares for FREE - Just find a open-source Project, and - Start your open-source Contribution

Google Chrome silently installs a 4 GB Gemini Nano AI model on your device. ❌ No consent ❌ No opt-in ❌ No real opt-out for regular users This isn’t a small experiment & it’s on billions of devices. ➡️ Read more: thatprivacyguy.com/blog/chrome-si… 😠 A tactic we've already seen with Gemini on Android. But at least you can disable it here, learn how: tuta.com/blog/how-to-di…

Google Chrome is quietly downloading a roughly 4 GB AI model to many users’ computers without clear upfront consent. The file, called weights.bin, is part of Google’s Gemini Nano on-device language model and lands in the browser’s user data folder under OptGuideOnDeviceModel. It powers built-in AI tools such as “Help me write,” smarter tab suggestions, on-device scam detection, and page summarization. The download triggers automatically for devices meeting minimum hardware requirements, and Chrome often replaces the files if deleted. While the model processes data locally, installation happens in the background with minimal notification. The scale is noteworthy. Hundreds of millions or billions of installations add up to thousands of tonnes of carbon emissions globally from data transfer, even though each is a one-time event. To prevent or remove it, go to chrome://flags, disable the entries for the optimization guide on-device model and Prompt API, restart the browser, and manually delete the folder.

been working on a TTS model that's better than elevenlabs and runs locally (for free) on Apple Silicon it's a fine tune of qwen3-tts with American voices that I've hand crafted to sound amazing some initial data: - TTFA 140ms - 2x real time factor - 0.6B at 6bit - duplex streaming - python & swift inference optimizing inference has been so fun honestly. the package will ship with an inference server thats so much better than anything on the market today. works on any M-series Mac with an insane 140ms latency. you can stream in token by token (made specifically for LLM use cases) and stream audio out as well. the server carries over prosody between sentences. can't wait to share more

Google Chrome silently installs a 4 GB AI model on your device. > No consent dialog. No opt-out UI. Re-installs itself if the user removes it manually. That is the true definition of malware.

Enjoy your life on @AmazingIslands #threejs #amazingislands #fortnite #nodejs #minecraft #bunjs #love #game #free #blender #3dtool @MoonriseEngine

Offline vibe coding with Gemma 4. This open-source Electron app runs entirely on your Mac via Apple's MLX framework. No Wi-Fi required.

Welcome @anthropicai as a Corporate Patron sponsor of the Blender Foundation’s Development Fund! This support will be dedicated towards general Blender core development. More details at blender.org/press/anthropi… #b3d #DevFund

Qwen 35B running locally on a MacBook. No lag. No API. No limits. Atomic Chat uses Google's TurboQuant under the hood 1,000+ open source models, fully offline, zero setup. LOCAL AI IS RUNNING OUT OF EXCUSES TO NOT USE IT.

We’re on the verge of interactive, real-time, photorealistic video generation with what are called World Models. These require a fair amount of expensive compute, but costs will come down over time. Today, these interactions are essentially real-time dreams. They lack the persistence and shared logic that turns a video into a multiplayer game. Or concert. Or classroom. Or the holodeck. We believe the ultimate architecture for gaming (and the holodeck) is Roblox Reality. It’s a hybrid architecture that marries the structured data and logic of the Roblox Engine and Roblox Cloud with the generative power of Video World Models. The Roblox Engine provides the underlying synchronized ground truth—the score, physics, multiplayer sync, etc.—while our video model acts as a Super Upsampler to layer on photorealistic detail. We believe this will ultimately remove barriers to high-fidelity creation, allowing a team of three people to build a narrative-driven, photorealistic masterpiece in a single week. This is an early look at turning solitary AI dreams into a social, playable reality. about.roblox.com/newsroom/2026/…

Microsoft Presents "TRELLIS.2": An Open-Source, 4B-Parameter, Image-to-3D Model producing up to 1536³ PBR textured assets. Built On Native 3D VAES With 16× Spatial compression, delivering efficient, scalable, high-fidelity asset generation. Ngl, pretty cool!

If I'm not mistaken, @elonmusk said that superior software architecture is AI's most underrated VECTOR. Software innovation, and their adaptive AI architectures, can/will redesign an entire CAPEX equation. From more compute with less energy, to new CHIPS, whether for general purpose or specific applications. Broadly speaking, I wonder if breakthroughs in AI Software will destroy today's standard CAPEX models - more so than having an edge on High Performance Chips. If true, Software poses as a severe threat to the "Brute Force" oriented, HIGH CAPEX AI Models. Is Software the Hyperscalers' "Achilles Hell"??? And if so, who's in line to become its first victims??? As we enter in the geopolitical aftermath of structural energy and supply chain disruptions... Data Center cancellations... Already high leverage ratios, with scarcer and less affordable financing alternatives... There's our recipe for Hyperscalers' Financial Distress and/or Wide-scale Consolidations.

Someone literally just made a desktop app that transforms 2D images into 3D models, running 100% LOCALLY on your own GPU 🤯 Meet Modly. Zero cloud, zero crazy API fees. Just drag, drop, and BOOM: instant 3D mesh 💥 Fully open-source.

Meet Kimi K2.6 agent - Video hero section, WebGL shaders, real backends. From one prompt. 🔹 Video hero sections - cinematic aesthetic, auto-composited 🔹 WebGL shader animations - native GLSL / WGSL, liquid metal, caustics, raymarching 🔹 Motion design - GSAP + Framer Motion 🔹 Backend database: Kimi wires up auth + database + backend in one pass. 🔹 Website stack - React 19 + TypeScript + Vite + Tailwind + shadcn/ui 🔹 3D w/ physically-based lighting - Three.js + React Three Fiber